How to Conduct a Patient Experience Data Self-Assessment

Question 1 and 2: Run Box and Calculation of Median 

Patient experience data can be well understood when run charts are employed. This not only creates a better analysis of patient experience but it also helps reveal deficiencies in service delivery. A run chart is a graph which details performance on determined aspects over time. When used, they help in formulation of appropriate goals that enhance improvements. To enhance the efficiency of run charts, a median is used as a reference point to determine deviation from expectations. Various rules are also employed to determine gaps that need to be addressed. In the run chart provided the median is 77. This is computed by arranging the values given in an ascending order and selecting the value that is centrally located. 

Question 3: Rules of a Run Chart 

The various rules that apply in a run chart determine the effectiveness of change applied to a given phenomenon and include shift, trend and astronomical point. Shift implies six or more points placed consecutively either all points lying above the median or below it. However, it is important to understand that points lying on the median are excluded. Trend on the other hand, implies five or more points placed consecutively in a descending or ascending order. Nevertheless, similar values do not form a trend and therefore only one value is selected. An astronomical point is a different value portrayed dramatically. Chart A contains both trend and shift and therefore it portrays improvement. Chart B on the other hand shows no change as there is no evidence of improvement or degradation. On the contrary, chart C portrays degradation due to presence of a negative trend. 

Question 4: Interpreting a New Month’s Value 

In interpreting the new month’s value of 76%, plotting it on a run chart is always justified. This is because a run chart would depict the effectiveness of interventions employed as unusual patterns are established. Looking at the willingness to recommend data in context of other elements of the survey is also justifiable. This is because a better insight is provided when related elements are analyzed in unison as compared to when a single element is analyzed. On the other hand, reiterating to staff the importance of continuous improvement and the financial implications of a relatively poor score is not justifiable. Analysis should rather be made to other related factors that may attribute to this poor score for example management practices employed. Similarly, investigation on what changed in the most recent month to establish the cause of decline is not justifiable. This is because such a move may only lead to wastage of resources if focus is only made to the previous month. The interpretation of the value in terms of the number of patients surveyed is justifiable. This is because such a value is a representative of the entire population. 

Question 5: Interpreting HCAHPS Percentile Tables 

The claim that a hospital averaging 83 scores lower more than 150 hospitals nationally is true in accordance to percentile rules. The claim that on average only 17% of patients surveyed did not experience top box communication is also true since 83% lies on the contrary. The claim that your hospital outperformed more than 75% of surveyed hospitals since it scores 78 may be false or difficult to tell. This is because your most recent data contains information on patients who were discharged after the survey period. This claim may also be hard to tell unless previous information is available on your performance. Additionally, the claim on hospital A being higher than hospital B cannot be told as additional information to further prove this claim is unavailable. The claim that out of 2 out of 3 patients contacted did not complete the survey is also true since non-response rate is taken into consideration. It is also true that comparing ratings and percentiles of specific departments makes sense as greater sensitivity is achieved. 

Question 6: Correlations 

It is true that correlation in each row cannot be larger than 1 but it may be zero or even negative one. This is because 1, 0 or -1 implies positive, non-linear and negative linear relationships consecutively. The claim that improving staff sensitivity to inconvenience will improve the overall rating of in-patient experience cannot be told as other factors should be put into consideration. It is true that including patients in decisions will yield higher overall ratings compared to reduction in noise in and around the hospital. This is because patient involvement displays proper utilization of available resources compared to noise reduction. The claim that room cleanliness is insignificant in overall rating may be false or difficult to tell. This is because a low correlation rating may imply clustering of values and hence the value should not be ignored.